COPYRIGHT 1999 American Society of Agricultural Engineers
"To improve our food supply, we must have some measure of how much iron we can expect to actually absorb," says Leon Kochian, head of the ARS U.S. Plant, Soil and Nutrition Laboratory in Ithaca, New York.
To that end, scientists at the Ithaca lab have developed an artificial gut that could enhance knowledge about the amount of iron available from food and food supplements. The research, led by human physiologist Raymond Glahn, resulted in an in vitro model that couples simulated food digestion with a human intestinal cell line. The model allows digestion to occur simultaneously with opportunity for nutrient uptake by Caco-2 cells. Caco-2 cells resemble the human intestinal epithelial cells that line the small intestine's inner surface and absorb nutrients from food, Kochian says.
To combine digestion with cell culture, Glahn needed a way to protect the Caco-2 cells lining the bottom surface of culture wells from digestive enzymes and micro-organisms. He achieved this protection by dividing the well into upper and lower chambers. As a separator, he used a dialysis membrane attached to a plastic insert that fits inside each half-dollar-sized culture well.
In the upper chamber -- directly on the membrane -- Glahn places a food sample and enzymes that digest it over a period of about three hours. The dialysis membrane prevents the digestive enzymes and microbes from reaching the Caco-2 cells in the lower chamber.
This method mimics the mucus layer that protects epithelial cells in the human digestive tract. Nutrients and minerals pass through the membrane to the Caco-2 cells. To determine how much of the food's iron is available to the Caco-2 cells, Glahn measures the amount of ferritin -- an iron storage protein -- in the cells.
Glahn's model is a fast, inexpensive method to determine the relative availability of iron from different foods or crop varieties used in the same food, Kochian says. For example, the model gut could screen dozens of rice varieties for iron availability.
So far, Glahn and co-workers have used the system to investigate iron availability in rice cereal, infant formulas and iron supplements. The model could have applications for studying staples such as corn, pharmaceutical iron preparations, food supplements, wheat and beans, and baby foods.
Glahn has applied for a U.S. patent on the model gut and is negotiating a license with a private laboratory to produce an iron bioavailability kit. The research team continues to apply, adapt and refine the model, which Glahn believes could someday be used to measure other micronutrients, such as vitamin, A, zinc, selenium and iodine.
Source Citation:"Artificial gut could help scientists digest more information about iron in foods." Resource: Engineering & Technology for a Sustainable World 6.11 (Nov 1999): 3. Academic OneFile. Gale. Salem High School. 22 Sept. 2009
<http://find.galegroup.com/gps/start.do?prodId=IPS>.
Characteristics of diet among a culturally diverse group of low-income pregnant women.
**//Journal of the//** American Dietetic Association 90.n4 (April 1990): p.p543(7). (4589 words)
Abstract:
During pregnancy, nutrition has an important impact on the health of both mother and child. Research concerning the food choices of pregnant women is useful for developing effective methods of nutrition counseling for this group. The diet quality of 335 low-income pregnant women was studied. The subject group was ethnically diverse and included mothers participating in the WIC (Women, Infants and Children) Program, which provides free nutritious foods to women with demonstrated need. Based on interviews and questionnaires, the subjects' intakes of protein and vitamin C were found to be more than adequate, but iron intake was poor. Calcium consumption was above the recommended level for white women, but just below it for non-white women. Ethnic background was associated with differences in the intake of several other nutrients. The diets of WIC participants were better in terms of calcium, iron, protein and vitamin B-6 than the diets of women who were not enrolled in WIC. Women who were employed had more nutritious diets than those who were unemployed, as did non-smokers versus smokers. Overall, this group of low-income pregnant women met or almost met their requirements for several important nutrients. The results indicate that when counseling pregnant women, emphasis should be placed on choosing foods that are nutrient-dense (high in nutrients on a per calorie basis), while less attention should be given to variety in the diet as a means of assuring nutritional adequacy. (Consumer Summary produced by Reliance Medical Information, Inc.)
Full Text :
COPYRIGHT The American Dietetic Association 1990 Abstract This study reports on diet quality and variety in food selection among a culturally diverse group of 335 low-income pregnant Massachusetts women. The index for Nutritional Quality (INQ), which is the observed nutrient density divided by the recommended nutrient density was estimated for seven nutrients from data reported on food frequency questionnaires and diet recalls. Mean INQs for protein and vitamin C were above the recommended level of 1.0, whereas INQ for iron was 0.6 or less. Mean INQ for calcium was 1.2 for whites and between 0.9 and 1.0 for non-whites. When comparisons were made among ethnic groups by site of prenatal care, significant differences in INQ were found for all nutrients except iron and vitamin B-6. WIC participants had higher diet quality for protein, calcium, iron, and vitamin B-6 than did non-participants. Employed women had higher protein and zinc INQs than did unemployed women, and non-smokers had higher iron and vitamin B-6 INQs than did smokers. Variety was not significantly correlated with diet quality but differed (p<.001) among site-specific ethnic groups. We conclude that there is a need for investigation of factors influencing dietary practices that are associated with sites of prenatal care. J Am Diet Assoc 90:543-549, 1990. Introduction
Knowledge of characteristics of prenatal food and nutrient use can assist in the development of improved approaches to dietary guidance. Detailed analyses based on nationwide surveys generally omit data collected from pregnant women because of small numbers. For example, the United States Department of Agriculture's 1977 to 1978 Nationwide Food Consumption Survey included only three tables reporting nutrient intake for 288 pregnant women out of a sample of 36,255 individuals (1). Thus, data from less representative studies of pregnant women may serve a useful role in examining issues related to prenatal dietary guidance.
Recommendations for food intake during pregnancy aim to guide women in making choices to achieve the Recommended Dietary Allowances (RDAs) (2) for nutrient intake. Hegsted (3) pointed out problems with use of the RDAs as standards for the general population. in their stead, he proposed use of standards based on nutrient densities, that is, nutrient intake per 1,000 kcal (4). Windham et al. (5,6) and Hansen et al. (7,8) also provided evidence supporting use of a nutrient density approach in nutrition education. Hansen (9) proposed an index for Nutritional Quality (INQ) to indicate diet quality in relation to RDAs. Most prenatal studies have reported nutrient intake only in absolute amounts.
Food guides for pregnancy recommend consumption of a minimum number of servings from at least four food groups and suggest choice of a variety of foods from within these groups. These recommendations are designed to promote a consistently adequate nutrient intake while allowing flexibility Variety is often promoted as a means of assuring adequate intake of all essential nutrients. Recently, however, questions have been raised regarding the usefulness of this concept (10, 11).
A study (I 2) to test the performance of a prenatal food frequency questionnaire (PFFQ) provided an opportunity to collect data about nutrient intake and food use over the preceding 4 weeks by a culturally diverse group of low-income women. In the present article, we examine data obtained via PFFQs or diet recalls with regard to diet quality and variety in food selection. Nutrients studied include protein, calcium, iron, zinc, total vitamin A (from both plant and animal sources), and vitamins B-6 and C.
We also compare our results with selected data from nationwide surveys and a recent study of pregnant women and explore associations of nutrient intake and food use with site of prenatal care, ethnic background, educational attainment, presence or absence of smoking, participation in federal food programs, employment, and economic status. Although the study provided data about use of vitamin/mineral supplements, those results are not included in this report. Methods
The sample population
Pregnant women were recruited from three Massachusetts health centers serving low-income clientele primarily, as described in detail elsewhere (12). Estimates obtained using the PFFQ for nutrients other than vitamin A were found to correlate well with those obtained from sets of three 24-hour diet recalls. All (364) recruited women were asked to provide demographic data and to complete a food frequency questionnaire; all gave the investigator permission to abstract selected information from the medical record. Of the women recruited, 344 (94.5%) provided data used in this report. Most (>85%) of these women were recruited on a follow-up prenatal visit after their first trimester of pregnancy Persistent attempts were made to contact a randomly drawn subsample of 160 women for diet recall interviews. Sets of three 24-hour diet recalls were obtained from 96 (60%) of them.
Of the total sample, 43% was white, 25% black, and 32% "other" (almost entirely Hispanics of Puerto Rican descent). Sixty-three percent of the women had incomes at or below the federal poverty level; most of the remainder had incomes of less than 200% of the poverty level. About two-thirds were unmarried. The women ranged in age from 14 to 43 years old; nearly 90% were at least 18 years old. The highest level of school completed ranged from sixth grade to graduation from college; 42% had not completed high school. Fifty-six percent reported that they had received benefits from the Supplemental Food Program for Women, Infants, and Children (WIC) in the past 4 weeks. A majority (65%) reported that they were not currently smoking cigarettes, whereas under 3% smoked more than one pack daily The sample of women who provided diet recalls was not entirely representative of the populations served by the clinics because it included smaller percentages of non-whites and young teens.
Dietary data collection methods
English and Spanish versions of the PFFQ were provided. The English version of the PFFQ included 90 food items and the Spanish version included 85, of which 79 were identical on the two versions. Subjects checked average frequency of food use over the past 4 weeks in one of nine columns ranging from "never" to 6 or more times/ day." No English and five Spanish questionnaires were excluded from the analysis because of failure to meet preset standards for technical errors (i.e., they contained >12 unmarked and/or doubly marked items). All other completed questionnaires were included in the analyses, regardless of language, unless otherwise specified. Mean daily nutrient intakes from the PFFQs and 3-day diet recalls were obtained as described elsewhere (12).
Statistical methods
The Statistical Package for the Social Sciences (SPSS/PC +) software package (1 3) was used for statistical computations. Nutrient data were transformed to logarithms to improve normality
Nutrient density was calculated using the 1989 RDAs (2) and the following formula: !!! BEGIN TABLE
nutrient intake
nutrient density = X 1,000.
caloric intake !!! END TABLE
The following formula was used to calculate INQs for pregnant women (9): !!! BEGIN TABLE
nutrient density of food consumed
INQ = _
RDA for pregnant woman/2,500 kcal !!! END TABLE
For vitamin A, 4,000 IU was used to approximate 800 RE from a mixed diet (14). Nutrient values from 328 PFFQs and 95 diet recall sets were used for INQ calculations.
There were strong associations between site of prenatal care and ethnic background in this study Site 1 served an approximately equal mix of white and Hispanic women, site 2 served mainly white women, and site 3 served mainly black women. Therefore, women were categorized into four groups on the basis of those two factors. The groupings used were white--site 1, Hispanic--site 1, white--site 2, and black--site 3. The small number of women belonging to other ethnic groups at each site were removed from this part of the analysis, decreasing total group size by about 8%.
Analysis of variance was used to test the hypotheses that there are no differences in INQ by site-specific ethnic groups, level of educational attainment high school graduate or not), participation in WIC, smoking status, and employment status. Scheffe's multiple comparison test was used when significant differences were found for the site-ethnicity factor.
Using diet recall data, we assessed the effect on nutrient quality of use of ready-to-eat cereal, many brands of which are highly fortified with vitamins and minerals. We applied Student's independent t-test to test the hypothesis that nutrient intake did not differ by reported use (at least once in three days) versus non-use of ready-to-eat cereal. For comparative purposes, we computed mean nutrient intake of cereal users excluding the nutrient contribution of cereal.
As its base, the variety score used 68 foods belonging to the dairy, meat, fruit and vegetable, and grains groups from the PFFQ food item list. Foods contributing mainly energy, such as cookies and soft drinks, were omitted. The variety score was defined as the number of the 68 foods used on a regular basis (at least once per week) in the past 4 weeks. Spanish versions of the PFFQ were excluded from this part of the analysis. Pearson product-moment correlation coefficients were computed to examine associations between variety score and nutrient density score.
To examine seasonal effects on variety, results obtained from site 1 during late summer were compared with results obtained in the spring, using Student's independent t-test. Similarly, results from site 3 were compared for late fall and late spring. Analysis of variance was used to examine associations of variety scores with site-specific ethnic group, smoking status, educational attainment, economic status (above versus at or below 100% of the federal poverty level, as determined from Medicaid insurance coverage), and participation in WIC or Food Stamp programs. Results
Quality of nutrient intake from food
An INQ of 1.0 indicates that the amount of nutrient per 1,000 kcal equals the recommended level for that nutrient. An INQ of less than 1.0 represents lower than recommended nutrient density Mean INQ for protein and vitamins B-6 and C exceeded 1.0 for all subgroups; calcium INQ varied somewhat by data collection method and by subgroup, but generally was close to 1.0. INQs for vitamin A were not compared because previous analysis had indicated that vitamin A values, as estimated from the PFFQ, were not correlated with values obtained from diet recall sets and were implausibly high on average (i.e., mean vitamin A intake exceeded 17,000 IU/day). Mean iron INQ for all subgroups was consistently below 0.7; the INQ for zinc was also low for each ethnic subgroup, falling between 0.7 and 0.9.
When analysis of variance was used to compare INQ by site-specific ethnic group and other factors, the educational variable was consistently unrelated to nutrient quality and was therefore discarded from the model. When PFFQs were used as the data source, site-specific ethnic group was a significant main effect for INQs for protein, calcium, zinc, and vitamins A and C. The direction of the effect was not consistent among the nutrients. For example, blacks at site 3 had a higher vitamin C INQ than did whites at site 1, whereas the reverse was true for the calcium INQ. Compared with nonparticipants, WIC participants had significantly higher INQs for protein, calcium, iron, and vitamin B-6, but the magnitude of the difference was small. Employed women had significantly higher protein and zinc INQs than did unemployed women. Current smokers ([is greater than or equal to] 1 cigarette daily) had INQs that were lower than those of non-smokers for iron and vitamin B-6. When the diet recall data were used for calculating INQs, site-specific ethnic group remained a significant main effect for calcium. No other notable differences were detected in the smaller (no. = 90) diet recall data set. There was virtually no difference in results whether or not data from the Spanish version of the PFFQ were included in the analysis.
in Table 3, diet quality, as measured by nutrient density from diet recall data, is compared with that reported by Endres et al. (15) for pregnant teens enrolled in WIC and by young women in the Continuing Survey of Food Intakes by Individuals (16,17). Differences in results for the two studies of pregnant women were inconsistent. Mean diet quality for calcium, iron, and vitamins B-6 and C was somewhat better for pregnant women in our study than was reported for a representative sample of low-income women (15), as well as for a representative survey of women of all income levels (17). Mean values for protein, although lower for our sample, are still higher than the recommended level.
Another way to examine diet quality is to determine the percentage of women reporting consumption of protective foods"--foods that are rich sources of vitamins, minerals, or protein--in a given period of time and the mean grams per day. Table 4 lists the protective foods" fluid milk, cheese, citrus fruit juice, ready-to-eat cereal, meat, dark green vegetables, and deep yellow vegetables (with composition of these food categories as specified in the USDA Continuing Survey of Food Intake [16]). All foods in the WIC package were used more frequently and in greater amounts by the pregnant women.
PFFQ results indicated that 92% of the women ate ready-to-eat cereal at least once a month; 67% of the women ate cereal at least once during the 3 days for which they provided diet recalls. The latter cereal users had higher mean intake of iron, zinc, and vitamins A, B-6, and C than did women who reported no cereal use (Table 5). Nutrient density for cereal users was significantly higher as well, for all of the same nutrients except vitamin C. The mean percent contribution of ready-to-eat cereal to total intake of the above-named nutrients was substantial (e.g., more than 30% of the total iron and vitamin B-6 intake).
Variety in food selection
Twelve foods were reportedly eaten at least two to four times weekly by 50% or more of the respondents (Table 6). This group of foods includes the five major components of the WIC package--milk, eggs, juice, cheese, and cereal.
Regardless of whether diet recall or PFFQ nutrient data were used, correlations between nutrient density scores and variety scores were low, ranging between - 0.17 and 0.26. Vitamin A density from the PFFQ was an exception (r=.36, two-tailed p =.01).
Mean variety score was 32.8 [+ or -] 11.3 (standard deviation), and the range was 2 to 66, before women were excluded to form site-specific ethnic groups. Mean variety score did not change noticeably after the exclusions. The women reporting use of only a few foods often commented that they could not tolerate many foods at that time. No significant differences in variety scores were found in the two comparisons by season.
There were statistically significant differences in variety score by site-specific ethnic groups (p<.001), but not by educational level or by interaction terms. Both the white and Hispanic women at site I had significantly lower variety scores than the groups at the other two sites (p = .005 for whites, p = .05 for Hispanics, Scheffe's multiple comparison test). Further analysis showed no significant differences by economic status, use of food stamps, or WIC participation, either as individual factors or combined with site-specific ethnic group. Discussion
Diet quality
To achieve recommended diet quality during pregnancy, a woman must ingest far more iron and zinc per 1,000 kcal than is characteristic of the general U.S. population. The WIC program attempts to help eligible low-income women improve the nutrient density of their diets by making available a package of foods high in protein and selected vitamins and minerals. WIC-approved cereals, for example, provide at least 28 mg iron per 100 mg dry cereal--a very high level achieved by fortification. Consuming all allotted items in a complete "WIC package' (18) practically assures achieving recommended diet quality for vitamin C, protein, and calcium, and makes a substantial contribution of iron, zinc, vitamins A and B-6, and several other nutrients not covered in this study
Other investigators have reported lower mean daily calcium intakes for blacks and "others" than for whites (16,19,20). Although the same relationship was observed in this study the calcium INQ for non-whites approached the target level, suggesting that many of these women had increased their consumption of calcium-rich foods.
It is useful to consider site of prenatal care as well as ethnic origin when comparing INQs because location may affect eating practices or dietary guidance and its effectiveness. Although significant differences in INQ tended to reflect published reports of ethnic differences in nutrient intake, it is notable that for some nutrients (i.e., protein, zinc, and vitamin A) means for two different cultural groups were closer in value than the means for the two white groups from different sites.
The lack of a clear association between educational level and nutrient intake has also been reported by Rush et al. (20) in their national evaluation of the WIC program, but this finding provides no information regarding the value of prenatal nutrition education efforts.
A large percentage of the low-income women in our study reported diets that were reasonably high in quality Their relatively frequent use of milk, cheese, and ready-to-eat cereal was probably linked to nutrition education and the supplemental WIC foods. Although INQs for zinc and iron appeared low, nutrient densities were in line with reports on absolute intake of those nutrients by the general population (I 6,21). The subjects' low INQs for those two nutrients are not necessarily indicative of poor food choices.
Variety in food selection
Because dietary guidance nearly always includes the recommendation to eat a variety of foods from the four food groups), several studies have been conducted to examine associations of variety of food intake with diet quality, biochemical measures of nutritional status, or demographic factors (10,11,22,23). We examined variety in relation to INQ (or nutrient density) rather than in relation to total nutrient intake. Smiciklas-Wright et al. (22) and Randall et al. (10) noted a strong association of total variety with total food use. Krebs-Smith et al. (11) controlled for total food use during analysis.
Although our methods differed somewhat from those used by other investigators, the lack of a clear association between the number of different "nutritious" foods consumed and nutrient quality is generally consistent with results from other studies (10,11,21,22).
Because of the unbalanced distribution of ethnic groups at the three sites, it was impossible to clearly separate ethnic group from site. The lower mean variety scores for whites at site 1 compared with those at site 2 suggest that factors such as availability, price, or local custom may have had a stronger influence on the variety score than did ethnic background. The similarity between variety scores for whites and Hispanics at site 1 supports that conclusion. Factors that contributed to variety of food use could not be determined from available data. Implications
Although our study design did not allow for conclusions regarding effectiveness of nutrition intervention in prenatal care, the data indicated that on average these low-income women in prenatal care achieved or approached recommended diet quality (INQs) for several of the nutrients studied. Furthermore, both WIC participation and nonsmoking status were associated with higher mean intake of some nutrients. Although INQs for iron and zinc were lower than recommended, this is typical of U.S. diets.
On the basis of our findings, we suggest increased attention to factors associated with site of prenatal care in relation to food and nutrient use of patients. We caution health care providers working with low-income women against placing heavy emphasis on variety as a means of improving diet quality
Complete Chicken Wing Dissection Lab Report (see previous page for links).
Play the Body System Matching Game (see previous page).
Summarize first article: "Artificial Gut..."
Read second article and answer the following questions:
Artificial gut could help scientists digest more information about iron in foods.(Brief Article).
**//Resource: Engineering & Technology for a Sustainable World//** 6.11 (Nov 1999): p.3. (428 words)
Full Text :
COPYRIGHT 1999 American Society of Agricultural Engineers"To improve our food supply, we must have some measure of how much iron we can expect to actually absorb," says Leon Kochian, head of the ARS U.S. Plant, Soil and Nutrition Laboratory in Ithaca, New York.
To that end, scientists at the Ithaca lab have developed an artificial gut that could enhance knowledge about the amount of iron available from food and food supplements. The research, led by human physiologist Raymond Glahn, resulted in an in vitro model that couples simulated food digestion with a human intestinal cell line. The model allows digestion to occur simultaneously with opportunity for nutrient uptake by Caco-2 cells. Caco-2 cells resemble the human intestinal epithelial cells that line the small intestine's inner surface and absorb nutrients from food, Kochian says.
To combine digestion with cell culture, Glahn needed a way to protect the Caco-2 cells lining the bottom surface of culture wells from digestive enzymes and micro-organisms. He achieved this protection by dividing the well into upper and lower chambers. As a separator, he used a dialysis membrane attached to a plastic insert that fits inside each half-dollar-sized culture well.
In the upper chamber -- directly on the membrane -- Glahn places a food sample and enzymes that digest it over a period of about three hours. The dialysis membrane prevents the digestive enzymes and microbes from reaching the Caco-2 cells in the lower chamber.
This method mimics the mucus layer that protects epithelial cells in the human digestive tract. Nutrients and minerals pass through the membrane to the Caco-2 cells. To determine how much of the food's iron is available to the Caco-2 cells, Glahn measures the amount of ferritin -- an iron storage protein -- in the cells.
Glahn's model is a fast, inexpensive method to determine the relative availability of iron from different foods or crop varieties used in the same food, Kochian says. For example, the model gut could screen dozens of rice varieties for iron availability.
So far, Glahn and co-workers have used the system to investigate iron availability in rice cereal, infant formulas and iron supplements. The model could have applications for studying staples such as corn, pharmaceutical iron preparations, food supplements, wheat and beans, and baby foods.
Glahn has applied for a U.S. patent on the model gut and is negotiating a license with a private laboratory to produce an iron bioavailability kit. The research team continues to apply, adapt and refine the model, which Glahn believes could someday be used to measure other micronutrients, such as vitamin, A, zinc, selenium and iodine.
Source Citation:"Artificial gut could help scientists digest more information about iron in foods." Resource: Engineering & Technology for a Sustainable World 6.11 (Nov 1999): 3. Academic OneFile. Gale. Salem High School. 22 Sept. 2009
<http://find.galegroup.com/gps/start.do?prodId=IPS>.
Characteristics of diet among a culturally diverse group of low-income pregnant women.
**//Journal of the//** American Dietetic Association 90.n4 (April 1990): p.p543(7). (4589 words)
Abstract:
During pregnancy, nutrition has an important impact on the health of both mother and child. Research concerning the food choices of pregnant women is useful for developing effective methods of nutrition counseling for this group. The diet quality of 335 low-income pregnant women was studied. The subject group was ethnically diverse and included mothers participating in the WIC (Women, Infants and Children) Program, which provides free nutritious foods to women with demonstrated need. Based on interviews and questionnaires, the subjects' intakes of protein and vitamin C were found to be more than adequate, but iron intake was poor. Calcium consumption was above the recommended level for white women, but just below it for non-white women. Ethnic background was associated with differences in the intake of several other nutrients. The diets of WIC participants were better in terms of calcium, iron, protein and vitamin B-6 than the diets of women who were not enrolled in WIC. Women who were employed had more nutritious diets than those who were unemployed, as did non-smokers versus smokers. Overall, this group of low-income pregnant women met or almost met their requirements for several important nutrients. The results indicate that when counseling pregnant women, emphasis should be placed on choosing foods that are nutrient-dense (high in nutrients on a per calorie basis), while less attention should be given to variety in the diet as a means of assuring nutritional adequacy. (Consumer Summary produced by Reliance Medical Information, Inc.)Full Text :
COPYRIGHT The American Dietetic Association 1990Abstract This study reports on diet quality and variety in food selection among a culturally diverse group of 335 low-income pregnant Massachusetts women. The index for Nutritional Quality (INQ), which is the observed nutrient density divided by the recommended nutrient density was estimated for seven nutrients from data reported on food frequency questionnaires and diet recalls. Mean INQs for protein and vitamin C were above the recommended level of 1.0, whereas INQ for iron was 0.6 or less. Mean INQ for calcium was 1.2 for whites and between 0.9 and 1.0 for non-whites. When comparisons were made among ethnic groups by site of prenatal care, significant differences in INQ were found for all nutrients except iron and vitamin B-6. WIC participants had higher diet quality for protein, calcium, iron, and vitamin B-6 than did non-participants. Employed women had higher protein and zinc INQs than did unemployed women, and non-smokers had higher iron and vitamin B-6 INQs than did smokers. Variety was not significantly correlated with diet quality but differed (p<.001) among site-specific ethnic groups. We conclude that there is a need for investigation of factors influencing dietary practices that are associated with sites of prenatal care. J Am Diet Assoc 90:543-549, 1990.
Introduction
Knowledge of characteristics of prenatal food and nutrient use can assist in the development of improved approaches to dietary guidance. Detailed analyses based on nationwide surveys generally omit data collected from pregnant women because of small numbers. For example, the United States Department of Agriculture's 1977 to 1978 Nationwide Food Consumption Survey included only three tables reporting nutrient intake for 288 pregnant women out of a sample of 36,255 individuals (1). Thus, data from less representative studies of pregnant women may serve a useful role in examining issues related to prenatal dietary guidance.
Recommendations for food intake during pregnancy aim to guide women in making choices to achieve the Recommended Dietary Allowances (RDAs) (2) for nutrient intake. Hegsted (3) pointed out problems with use of the RDAs as standards for the general population. in their stead, he proposed use of standards based on nutrient densities, that is, nutrient intake per 1,000 kcal (4). Windham et al. (5,6) and Hansen et al. (7,8) also provided evidence supporting use of a nutrient density approach in nutrition education. Hansen (9) proposed an index for Nutritional Quality (INQ) to indicate diet quality in relation to RDAs. Most prenatal studies have reported nutrient intake only in absolute amounts.
Food guides for pregnancy recommend consumption of a minimum number of servings from at least four food groups and suggest choice of a variety of foods from within these groups. These recommendations are designed to promote a consistently adequate nutrient intake while allowing flexibility Variety is often promoted as a means of assuring adequate intake of all essential nutrients. Recently, however, questions have been raised regarding the usefulness of this concept (10, 11).
A study (I 2) to test the performance of a prenatal food frequency questionnaire (PFFQ) provided an opportunity to collect data about nutrient intake and food use over the preceding 4 weeks by a culturally diverse group of low-income women. In the present article, we examine data obtained via PFFQs or diet recalls with regard to diet quality and variety in food selection. Nutrients studied include protein, calcium, iron, zinc, total vitamin A (from both plant and animal sources), and vitamins B-6 and C.
We also compare our results with selected data from nationwide surveys and a recent study of pregnant women and explore associations of nutrient intake and food use with site of prenatal care, ethnic background, educational attainment, presence or absence of smoking, participation in federal food programs, employment, and economic status. Although the study provided data about use of vitamin/mineral supplements, those results are not included in this report.
Methods
The sample population
Pregnant women were recruited from three Massachusetts health centers serving low-income clientele primarily, as described in detail elsewhere (12). Estimates obtained using the PFFQ for nutrients other than vitamin A were found to correlate well with those obtained from sets of three 24-hour diet recalls. All (364) recruited women were asked to provide demographic data and to complete a food frequency questionnaire; all gave the investigator permission to abstract selected information from the medical record. Of the women recruited, 344 (94.5%) provided data used in this report. Most (>85%) of these women were recruited on a follow-up prenatal visit after their first trimester of pregnancy Persistent attempts were made to contact a randomly drawn subsample of 160 women for diet recall interviews. Sets of three 24-hour diet recalls were obtained from 96 (60%) of them.
Of the total sample, 43% was white, 25% black, and 32% "other" (almost entirely Hispanics of Puerto Rican descent). Sixty-three percent of the women had incomes at or below the federal poverty level; most of the remainder had incomes of less than 200% of the poverty level. About two-thirds were unmarried. The women ranged in age from 14 to 43 years old; nearly 90% were at least 18 years old. The highest level of school completed ranged from sixth grade to graduation from college; 42% had not completed high school. Fifty-six percent reported that they had received benefits from the Supplemental Food Program for Women, Infants, and Children (WIC) in the past 4 weeks. A majority (65%) reported that they were not currently smoking cigarettes, whereas under 3% smoked more than one pack daily The sample of women who provided diet recalls was not entirely representative of the populations served by the clinics because it included smaller percentages of non-whites and young teens.
Dietary data collection methods
English and Spanish versions of the PFFQ were provided. The English version of the PFFQ included 90 food items and the Spanish version included 85, of which 79 were identical on the two versions. Subjects checked average frequency of food use over the past 4 weeks in one of nine columns ranging from "never" to 6 or more times/ day." No English and five Spanish questionnaires were excluded from the analysis because of failure to meet preset standards for technical errors (i.e., they contained >12 unmarked and/or doubly marked items). All other completed questionnaires were included in the analyses, regardless of language, unless otherwise specified. Mean daily nutrient intakes from the PFFQs and 3-day diet recalls were obtained as described elsewhere (12).
Statistical methods
The Statistical Package for the Social Sciences (SPSS/PC +) software package (1 3) was used for statistical computations. Nutrient data were transformed to logarithms to improve normality
Nutrient density was calculated using the 1989 RDAs (2) and the following formula: !!! BEGIN TABLE
nutrient intake
nutrient density = X 1,000.
caloric intake !!! END TABLE
The following formula was used to calculate INQs for pregnant women (9): !!! BEGIN TABLE
nutrient density of food consumed
INQ = _
RDA for pregnant woman/2,500 kcal !!! END TABLE
For vitamin A, 4,000 IU was used to approximate 800 RE from a mixed diet (14). Nutrient values from 328 PFFQs and 95 diet recall sets were used for INQ calculations.
There were strong associations between site of prenatal care and ethnic background in this study Site 1 served an approximately equal mix of white and Hispanic women, site 2 served mainly white women, and site 3 served mainly black women. Therefore, women were categorized into four groups on the basis of those two factors. The groupings used were white--site 1, Hispanic--site 1, white--site 2, and black--site 3. The small number of women belonging to other ethnic groups at each site were removed from this part of the analysis, decreasing total group size by about 8%.
Analysis of variance was used to test the hypotheses that there are no differences in INQ by site-specific ethnic groups, level of educational attainment high school graduate or not), participation in WIC, smoking status, and employment status. Scheffe's multiple comparison test was used when significant differences were found for the site-ethnicity factor.
Using diet recall data, we assessed the effect on nutrient quality of use of ready-to-eat cereal, many brands of which are highly fortified with vitamins and minerals. We applied Student's independent t-test to test the hypothesis that nutrient intake did not differ by reported use (at least once in three days) versus non-use of ready-to-eat cereal. For comparative purposes, we computed mean nutrient intake of cereal users excluding the nutrient contribution of cereal.
As its base, the variety score used 68 foods belonging to the dairy, meat, fruit and vegetable, and grains groups from the PFFQ food item list. Foods contributing mainly energy, such as cookies and soft drinks, were omitted. The variety score was defined as the number of the 68 foods used on a regular basis (at least once per week) in the past 4 weeks. Spanish versions of the PFFQ were excluded from this part of the analysis. Pearson product-moment correlation coefficients were computed to examine associations between variety score and nutrient density score.
To examine seasonal effects on variety, results obtained from site 1 during late summer were compared with results obtained in the spring, using Student's independent t-test. Similarly, results from site 3 were compared for late fall and late spring. Analysis of variance was used to examine associations of variety scores with site-specific ethnic group, smoking status, educational attainment, economic status (above versus at or below 100% of the federal poverty level, as determined from Medicaid insurance coverage), and participation in WIC or Food Stamp programs.
Results
Quality of nutrient intake from food
An INQ of 1.0 indicates that the amount of nutrient per 1,000 kcal equals the recommended level for that nutrient. An INQ of less than 1.0 represents lower than recommended nutrient density Mean INQ for protein and vitamins B-6 and C exceeded 1.0 for all subgroups; calcium INQ varied somewhat by data collection method and by subgroup, but generally was close to 1.0. INQs for vitamin A were not compared because previous analysis had indicated that vitamin A values, as estimated from the PFFQ, were not correlated with values obtained from diet recall sets and were implausibly high on average (i.e., mean vitamin A intake exceeded 17,000 IU/day). Mean iron INQ for all subgroups was consistently below 0.7; the INQ for zinc was also low for each ethnic subgroup, falling between 0.7 and 0.9.
When analysis of variance was used to compare INQ by site-specific ethnic group and other factors, the educational variable was consistently unrelated to nutrient quality and was therefore discarded from the model. When PFFQs were used as the data source, site-specific ethnic group was a significant main effect for INQs for protein, calcium, zinc, and vitamins A and C. The direction of the effect was not consistent among the nutrients. For example, blacks at site 3 had a higher vitamin C INQ than did whites at site 1, whereas the reverse was true for the calcium INQ. Compared with nonparticipants, WIC participants had significantly higher INQs for protein, calcium, iron, and vitamin B-6, but the magnitude of the difference was small. Employed women had significantly higher protein and zinc INQs than did unemployed women. Current smokers ([is greater than or equal to] 1 cigarette daily) had INQs that were lower than those of non-smokers for iron and vitamin B-6. When the diet recall data were used for calculating INQs, site-specific ethnic group remained a significant main effect for calcium. No other notable differences were detected in the smaller (no. = 90) diet recall data set. There was virtually no difference in results whether or not data from the Spanish version of the PFFQ were included in the analysis.
in Table 3, diet quality, as measured by nutrient density from diet recall data, is compared with that reported by Endres et al. (15) for pregnant teens enrolled in WIC and by young women in the Continuing Survey of Food Intakes by Individuals (16,17). Differences in results for the two studies of pregnant women were inconsistent. Mean diet quality for calcium, iron, and vitamins B-6 and C was somewhat better for pregnant women in our study than was reported for a representative sample of low-income women (15), as well as for a representative survey of women of all income levels (17). Mean values for protein, although lower for our sample, are still higher than the recommended level.
Another way to examine diet quality is to determine the percentage of women reporting consumption of protective foods"--foods that are rich sources of vitamins, minerals, or protein--in a given period of time and the mean grams per day. Table 4 lists the protective foods" fluid milk, cheese, citrus fruit juice, ready-to-eat cereal, meat, dark green vegetables, and deep yellow vegetables (with composition of these food categories as specified in the USDA Continuing Survey of Food Intake [16]). All foods in the WIC package were used more frequently and in greater amounts by the pregnant women.
PFFQ results indicated that 92% of the women ate ready-to-eat cereal at least once a month; 67% of the women ate cereal at least once during the 3 days for which they provided diet recalls. The latter cereal users had higher mean intake of iron, zinc, and vitamins A, B-6, and C than did women who reported no cereal use (Table 5). Nutrient density for cereal users was significantly higher as well, for all of the same nutrients except vitamin C. The mean percent contribution of ready-to-eat cereal to total intake of the above-named nutrients was substantial (e.g., more than 30% of the total iron and vitamin B-6 intake).
Variety in food selection
Twelve foods were reportedly eaten at least two to four times weekly by 50% or more of the respondents (Table 6). This group of foods includes the five major components of the WIC package--milk, eggs, juice, cheese, and cereal.
Regardless of whether diet recall or PFFQ nutrient data were used, correlations between nutrient density scores and variety scores were low, ranging between - 0.17 and 0.26. Vitamin A density from the PFFQ was an exception (r=.36, two-tailed p =.01).
Mean variety score was 32.8 [+ or -] 11.3 (standard deviation), and the range was 2 to 66, before women were excluded to form site-specific ethnic groups. Mean variety score did not change noticeably after the exclusions. The women reporting use of only a few foods often commented that they could not tolerate many foods at that time. No significant differences in variety scores were found in the two comparisons by season.
There were statistically significant differences in variety score by site-specific ethnic groups (p<.001), but not by educational level or by interaction terms. Both the white and Hispanic women at site I had significantly lower variety scores than the groups at the other two sites (p = .005 for whites, p = .05 for Hispanics, Scheffe's multiple comparison test). Further analysis showed no significant differences by economic status, use of food stamps, or WIC participation, either as individual factors or combined with site-specific ethnic group. Discussion
Diet quality
To achieve recommended diet quality during pregnancy, a woman must ingest far more iron and zinc per 1,000 kcal than is characteristic of the general U.S. population. The WIC program attempts to help eligible low-income women improve the nutrient density of their diets by making available a package of foods high in protein and selected vitamins and minerals. WIC-approved cereals, for example, provide at least 28 mg iron per 100 mg dry cereal--a very high level achieved by fortification. Consuming all allotted items in a complete "WIC package' (18) practically assures achieving recommended diet quality for vitamin C, protein, and calcium, and makes a substantial contribution of iron, zinc, vitamins A and B-6, and several other nutrients not covered in this study
Other investigators have reported lower mean daily calcium intakes for blacks and "others" than for whites (16,19,20). Although the same relationship was observed in this study the calcium INQ for non-whites approached the target level, suggesting that many of these women had increased their consumption of calcium-rich foods.
It is useful to consider site of prenatal care as well as ethnic origin when comparing INQs because location may affect eating practices or dietary guidance and its effectiveness. Although significant differences in INQ tended to reflect published reports of ethnic differences in nutrient intake, it is notable that for some nutrients (i.e., protein, zinc, and vitamin A) means for two different cultural groups were closer in value than the means for the two white groups from different sites.
The lack of a clear association between educational level and nutrient intake has also been reported by Rush et al. (20) in their national evaluation of the WIC program, but this finding provides no information regarding the value of prenatal nutrition education efforts.
A large percentage of the low-income women in our study reported diets that were reasonably high in quality Their relatively frequent use of milk, cheese, and ready-to-eat cereal was probably linked to nutrition education and the supplemental WIC foods. Although INQs for zinc and iron appeared low, nutrient densities were in line with reports on absolute intake of those nutrients by the general population (I 6,21). The subjects' low INQs for those two nutrients are not necessarily indicative of poor food choices.
Variety in food selection
Because dietary guidance nearly always includes the recommendation to eat a variety of foods from the four food groups), several studies have been conducted to examine associations of variety of food intake with diet quality, biochemical measures of nutritional status, or demographic factors (10,11,22,23). We examined variety in relation to INQ (or nutrient density) rather than in relation to total nutrient intake. Smiciklas-Wright et al. (22) and Randall et al. (10) noted a strong association of total variety with total food use. Krebs-Smith et al. (11) controlled for total food use during analysis.
Although our methods differed somewhat from those used by other investigators, the lack of a clear association between the number of different "nutritious" foods consumed and nutrient quality is generally consistent with results from other studies (10,11,21,22).
Because of the unbalanced distribution of ethnic groups at the three sites, it was impossible to clearly separate ethnic group from site. The lower mean variety scores for whites at site 1 compared with those at site 2 suggest that factors such as availability, price, or local custom may have had a stronger influence on the variety score than did ethnic background. The similarity between variety scores for whites and Hispanics at site 1 supports that conclusion. Factors that contributed to variety of food use could not be determined from available data.
Implications
Although our study design did not allow for conclusions regarding effectiveness of nutrition intervention in prenatal care, the data indicated that on average these low-income women in prenatal care achieved or approached recommended diet quality (INQs) for several of the nutrients studied. Furthermore, both WIC participation and nonsmoking status were associated with higher mean intake of some nutrients. Although INQs for iron and zinc were lower than recommended, this is typical of U.S. diets.
On the basis of our findings, we suggest increased attention to factors associated with site of prenatal care in relation to food and nutrient use of patients. We caution health care providers working with low-income women against placing heavy emphasis on variety as a means of improving diet quality